[jira] [Commented] (HIVE-14844) Not able to create the Hive table with more than 700 columns

["Nirmalkumar (JIRA)" <ji...@apache.org>]

    [ https://issues.apache.org/jira/browse/HIVE-14844?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15525392#comment-15525392 ] 

Nirmalkumar commented on HIVE-14844:
------------------------------------

It seems like we are hitting a metastore limitation related to the Hive Metastore "SERDE_PARAMS.PARAM_VALUE" column limits to varchar 4000 our SERDEPROPERTIES is 5620byte.


> Not able to create the Hive table with more than 700 columns
> ------------------------------------------------------------
>
>                 Key: HIVE-14844
>                 URL: https://issues.apache.org/jira/browse/HIVE-14844
>             Project: Hive
>          Issue Type: Bug
>          Components: Metastore
>    Affects Versions: 1.1.0
>         Environment: MySQL
>            Reporter: Nirmalkumar
>
> We tried creating the Hive table with 700+ columns which leads to the below error:
> FAILED: Execution Error, return code 1 from org.apache.hadoop.hive.ql.exec.DDLTask. MetaException(message:javax.jdo.JDODataStoreException: Put request failed : INSERT INTO `SERDE_PARAMS` (`PARAM_VALUE`,`SERDE_ID`,`PARAM_KEY`) VALUES (?,?,?)
> we use hive 1.1.0 and below is the SERDE properties we have in our MySQL.
> mysql> desc SERDE_PARAMS;
> +-------------+---------------+------+-----+---------+-------+
> | Field       | Type          | Null | Key | Default | Extra |
> +-------------+---------------+------+-----+---------+-------+
> | SERDE_ID    | bigint(20)    | NO   | PRI | NULL    |       |
> | PARAM_KEY   | varchar(256)  | NO   | PRI | NULL    |       |
> | PARAM_VALUE | varchar(4000) | YES  |     | NULL    |       |
> +-------------+---------------+------+-----+---------+-------+
> As per Cloudera,
> Actually it's a known limitation and there are no patches for this. 
> To avoid this issue typically is increased the size of the field in the metastore, BUT this is not a supported Cloudera solution that can result in a break in a future Metastore schema upgrade (in case if we are evolving our CDH version to an upper one that will implement an hive version greater than the actual one).
> So could you please provide a fix for this?



--
This message was sent by Atlassian JIRA
(v6.3.4#6332)